Airport Connection Network Analysis
- Data set Description:
The routes & airport data are from www.openflights.org. The routes data was last updated in 2014, while the airport data was last updated in 2017.
#Loading the necessary libraries:
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)
library(igraph)##
## Attaching package: 'igraph'## The following objects are masked from 'package:dplyr':
##
## as_data_frame, groups, union## The following objects are masked from 'package:stats':
##
## decompose, spectrum## The following object is masked from 'package:base':
##
## unionlibrary(itertools)## Loading required package: iteratorslibrary(psych)##
## Attaching package: 'psych'## The following objects are masked from 'package:ggplot2':
##
## %+%, alphalibrary(rgexf)
library(ggrepel)
library(RgoogleMaps)
library(ggmap)## Google's Terms of Service: https://cloud.google.com/maps-platform/terms/.## Please cite ggmap if you use it! See citation("ggmap") for details.library(mapproj)## Loading required package: mapslibrary(sf)## Linking to GEOS 3.8.1, GDAL 3.1.4, PROJ 6.3.1library(OpenStreetMap)
library(devtools)## Loading required package: usethislibrary(DT)
library(plyr)## ------------------------------------------------------------------------------## You have loaded plyr after dplyr - this is likely to cause problems.
## If you need functions from both plyr and dplyr, please load plyr first, then dplyr:
## library(plyr); library(dplyr)## ------------------------------------------------------------------------------##
## Attaching package: 'plyr'## The following object is masked from 'package:maps':
##
## ozone## The following objects are masked from 'package:dplyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarizelibrary(geosphere) # For spatial methods
library(threejs) # threejs is used for 3-D interactive Earth Visualization
library(rworldmap) # For creating earth map## Loading required package: sp## ### Welcome to rworldmap ##### For a short introduction type : vignette('rworldmap')library(leaflet) # Leaflet for R provides functions to control and integrate Leaflet, a JavaScript library for interactive maps, within R.
library(rgeos) # Provides functions for handling operations on topologies.## rgeos version: 0.5-5, (SVN revision 640)
## GEOS runtime version: 3.8.1-CAPI-1.13.3
## Linking to sp version: 1.4-2
## Polygon checking: TRUElibrary(raster) # For raster image##
## Attaching package: 'raster'## The following object is masked from 'package:dplyr':
##
## selectlibrary(DT) # For creating interactive tables
library(ggplot2)
library(sp) # For Spatial processing of data
library(ggmap) # To reverse geocode Long/Lat
library(knitr) # TO enable 3-D visualization embedding in the HTML page
library(rglwidget)## The functions in the rglwidget package have been moved to rgl.library(rgl)##
## Attaching package: 'rgl'## The following object is masked from 'package:rgeos':
##
## triangulate## The following objects are masked from 'package:threejs':
##
## lines3d, points3dlibrary(sqldf)## Loading required package: gsubfn## Loading required package: proto## Loading required package: RSQLite1. Load data & Graph
1.1 Loading Data & Examine Dataframe
routes_url <- "https://gist.githubusercontent.com/hannahbhchou/8f79bddf4ad93a573ada0d10453fe7d5/raw/a3b2624b38579d0c450d76532031f3f47a269dec/routes.csv"
airport_url <- "https://gist.githubusercontent.com/hannahbhchou/5f59fb70e3d287c577af4b1d74a13cb5/raw/98ec7a19cbe39bd92857280fd8a02e80c9ea249f/airports.csv"
routes_df <- read.csv(routes_url, header = TRUE )
airport_df <- read.csv(airport_url, header = TRUE)head(routes_df)## airline airline.ID source.airport source.airport.id destination.airport
## 1 2B 410 AER 2965 KZN
## 2 2B 410 ASF 2966 KZN
## 3 2B 410 ASF 2966 MRV
## 4 2B 410 CEK 2968 KZN
## 5 2B 410 CEK 2968 OVB
## 6 2B 410 DME 4029 KZN
## destination.airport.id codeshare stops equipment
## 1 2990 0 CR2
## 2 2990 0 CR2
## 3 2962 0 CR2
## 4 2990 0 CR2
## 5 4078 0 CR2
## 6 2990 0 CR2head(airport_df)## Airport.ID Name City
## 1 1 Goroka Airport Goroka
## 2 2 Madang Airport Madang
## 3 3 Mount Hagen Kagamuga Airport Mount Hagen
## 4 4 Nadzab Airport Nadzab
## 5 5 Port Moresby Jacksons International Airport Port Moresby
## 6 6 Wewak International Airport Wewak
## Country IATA ICAO Latitude Longtitude Altitude Timezone DST
## 1 Papua New Guinea GKA AYGA -6.081690 145.392 5282 10 U
## 2 Papua New Guinea MAG AYMD -5.207080 145.789 20 10 U
## 3 Papua New Guinea HGU AYMH -5.826790 144.296 5388 10 U
## 4 Papua New Guinea LAE AYNZ -6.569803 146.726 239 10 U
## 5 Papua New Guinea POM AYPY -9.443380 147.220 146 10 U
## 6 Papua New Guinea WWK AYWK -3.583830 143.669 19 10 U
## Tz.database.time.zone Type Source
## 1 Pacific/Port_Moresby airport OurAirports
## 2 Pacific/Port_Moresby airport OurAirports
## 3 Pacific/Port_Moresby airport OurAirports
## 4 Pacific/Port_Moresby airport OurAirports
## 5 Pacific/Port_Moresby airport OurAirports
## 6 Pacific/Port_Moresby airport OurAirportsstr(routes_df)## 'data.frame': 67663 obs. of 9 variables:
## $ airline : chr "2B" "2B" "2B" "2B" ...
## $ airline.ID : chr "410" "410" "410" "410" ...
## $ source.airport : chr "AER" "ASF" "ASF" "CEK" ...
## $ source.airport.id : chr "2965" "2966" "2966" "2968" ...
## $ destination.airport : chr "KZN" "KZN" "MRV" "KZN" ...
## $ destination.airport.id: chr "2990" "2990" "2962" "2990" ...
## $ codeshare : chr "" "" "" "" ...
## $ stops : int 0 0 0 0 0 0 0 0 0 0 ...
## $ equipment : chr "CR2" "CR2" "CR2" "CR2" ...str(airport_df)## 'data.frame': 7698 obs. of 14 variables:
## $ Airport.ID : int 1 2 3 4 5 6 7 8 9 10 ...
## $ Name : chr "Goroka Airport" "Madang Airport" "Mount Hagen Kagamuga Airport" "Nadzab Airport" ...
## $ City : chr "Goroka" "Madang" "Mount Hagen" "Nadzab" ...
## $ Country : chr "Papua New Guinea" "Papua New Guinea" "Papua New Guinea" "Papua New Guinea" ...
## $ IATA : chr "GKA" "MAG" "HGU" "LAE" ...
## $ ICAO : chr "AYGA" "AYMD" "AYMH" "AYNZ" ...
## $ Latitude : num -6.08 -5.21 -5.83 -6.57 -9.44 ...
## $ Longtitude : num 145 146 144 147 147 ...
## $ Altitude : int 5282 20 5388 239 146 19 112 283 165 251 ...
## $ Timezone : chr "10" "10" "10" "10" ...
## $ DST : chr "U" "U" "U" "U" ...
## $ Tz.database.time.zone: chr "Pacific/Port_Moresby" "Pacific/Port_Moresby" "Pacific/Port_Moresby" "Pacific/Port_Moresby" ...
## $ Type : chr "airport" "airport" "airport" "airport" ...
## $ Source : chr "OurAirports" "OurAirports" "OurAirports" "OurAirports" ...#drop unnecessary columns
airport_drop_col <- c("ICAO","Altitude","Timezone","DST", "Tz.database.time.zone","Type", "Source")
routes_drop_col <- c("codeshare","stops","equipment")
airport_df <- airport_df %>% dplyr::select(-one_of(airport_drop_col))
routes_df <- routes_df %>% dplyr::select(-one_of(routes_drop_col))1.2 Graph from dataframe & Graph Attributes
routes_edges <- routes_df %>% dplyr::select("source.airport", "destination.airport")
g <- graph_from_data_frame(d = routes_edges, directed = TRUE)num_edge <- gsize(g)
num_vertex <- gorder(g)
print(paste("There are", num_edge, "edges."))## [1] "There are 67663 edges."print(paste("There are", num_vertex, "vertices."))## [1] "There are 3425 vertices."1.3 Initial Plotting
plot(g, vertex.label= NA, layout = layout_nicely(g))
Already we could see there nodes more on the outskirt, the lonely islands in terms of air traffic
2. Centrality Measures
We are using 3 centrality measures to evaluate the nodes of our graph.
#Run all measurements
degree_vec <- degree(g)
betweenness_vec <- betweenness(g)
closeness_vec <- closeness(g)## Warning in closeness(g): At centrality.c:2784 :closeness centrality is not well-
## defined for disconnected graphsin_degree_vec <- degree(g, mode = "in")
out_degree_vec <- degree(g, mode = "out")
eigen_vec <- eigen_centrality(g)$vector#Attaching measures to the airport_df
degree_df <- as.data.frame(as.table(degree_vec))
betweenness_df <- as.data.frame(as.table(betweenness_vec))
closeness_df <- as.data.frame(as.table(closeness_vec))
in_degree_df <- as.data.frame(as.table(in_degree_vec))
out_degree_df <- as.data.frame(as.table(out_degree_vec))
eigen_df <- as.data.frame(as.table(eigen_vec))
names(degree_df)[1] <- "id"
names(degree_df)[2] <- "degree"
names(betweenness_df)[1] <- "id"
names(betweenness_df)[2] <- "betweenness"
names(closeness_df)[1] <- "id"
names(closeness_df)[2] <- "closeness"
names(in_degree_df)[1] <- "id"
names(in_degree_df)[2] <- "in_degree"
names(out_degree_df)[1] <- "id"
names(out_degree_df)[2] <- "out_degree"
names(eigen_df)[1] <- "id"
names(eigen_df)[2] <- "eigenvector"
airport_df <- airport_df %>% left_join(degree_df, by = c("IATA" = "id")) %>%
left_join(in_degree_df, by = c("IATA" = "id")) %>%
left_join(out_degree_df, by = c("IATA" = "id")) %>%
left_join(betweenness_df, by = c("IATA" = "id")) %>%
left_join(closeness_df, by = c("IATA" = "id")) %>%
left_join(eigen_df, by = c("IATA" = "id"))
airport_df <- airport_df[complete.cases(airport_df),]2.1 Degree Centrality
2.1.1 Maximum & Minimum Degree
max_degree <- max(degree_vec)
min_degree <- min(degree_vec)
print(paste("Maximum degree is", max_degree, "degree."))## [1] "Maximum degree is 1826 degree."print(paste("Minimum degree is", min_degree, "degree."))## [1] "Minimum degree is 1 degree."max_in_degree <- max(in_degree_vec)
min_in_degree <- min(in_degree_vec)
print(paste("Maximum in degree is", max_in_degree, "degree, which means this airport receives flights from", max_in_degree, "destinations."))## [1] "Maximum in degree is 911 degree, which means this airport receives flights from 911 destinations."print(paste("Minimum degree is", min_in_degree, "degree, which means this airport doesn't receive any flights."))## [1] "Minimum degree is 0 degree, which means this airport doesn't receive any flights."max_out_degree <- max(out_degree_vec)
min_out_degree <- min(out_degree_vec)
print(paste("Maximum out degree is", max_out_degree, "degree, which means this airport receives flights from", max_out_degree, "destinations."))## [1] "Maximum out degree is 915 degree, which means this airport receives flights from 915 destinations."print(paste("Minimum degree is", min_out_degree, "degree, which means this airport doesn't have departing flights."))## [1] "Minimum degree is 0 degree, which means this airport doesn't have departing flights."2.1.2 Top 20 Degree Airport
top20_degree_df <- airport_df[order(airport_df$degree, decreasing = TRUE),][1:20,c("IATA", "Name", "Country", "City","degree")]
top20_degree_df## IATA Name Country
## 3483 ATL Hartsfield Jackson Atlanta International Airport United States
## 3631 ORD Chicago O'Hare International Airport United States
## 3171 PEK Beijing Capital International Airport China
## 503 LHR London Heathrow Airport United Kingdom
## 1347 CDG Charles de Gaulle International Airport France
## 337 FRA Frankfurt am Main Airport Germany
## 3286 LAX Los Angeles International Airport United States
## 3471 DFW Dallas Fort Worth International Airport United States
## 3598 JFK John F Kennedy International Airport United States
## 575 AMS Amsterdam Airport Schiphol Netherlands
## 3208 PVG Shanghai Pudong International Airport China
## 3125 SIN Singapore Changi Airport Singapore
## 1187 BCN Barcelona International Airport Spain
## 3726 ICN Incheon International Airport South Korea
## 3552 DEN Denver International Airport United States
## 3377 MIA Miami International Airport United States
## 343 MUC Munich Airport Germany
## 7630 IST Istanbul Airport Turkey
## 2101 DXB Dubai International Airport United Arab Emirates
## 2916 HKG Hong Kong International Airport Hong Kong
## City degree
## 3483 Atlanta 1826
## 3631 Chicago 1108
## 3171 Beijing 1069
## 503 London 1051
## 1347 Paris 1041
## 337 Frankfurt 990
## 3286 Los Angeles 990
## 3471 Dallas-Fort Worth 936
## 3598 New York 911
## 575 Amsterdam 903
## 3208 Shanghai 825
## 3125 Singapore 820
## 1187 Barcelona 783
## 3726 Seoul 740
## 3552 Denver 735
## 3377 Miami 734
## 343 Munich 728
## 7630 Istanbul 719
## 2101 Dubai 710
## 2916 Hong Kong 710#set ggplot theme
world_theme <- theme(panel.background = element_rect(fill = "lightblue",
colour = "lightblue"),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
# surpress legend
legend.position = "none",
axis.line=element_blank(),
axis.text.x=element_blank(),
axis.text.y=element_blank(),
axis.ticks=element_blank(),
axis.title.x=element_blank(),
axis.title.y=element_blank())#set lower bound for label showing
thres <-top20_degree_df[20, "degree"]
degree_plot <- ggplot(airport_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
stat_density2d(aes(fill = ..level.., alpha = I(.3)),
size = 1, bins = 5, data = airport_df,
geom = "polygon") +
geom_point(color="red", alpha = .2, size=airport_df$degree/150) +
# define color of density polygons
scale_fill_gradient(low = "grey50", high = "grey20") +
world_theme +
geom_text_repel(data = subset(airport_df, degree > thres), aes(x=Longtitude, y= Latitude, label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) + ggtitle("By Degree")
degree_plot
top20_in_degree_df <- airport_df[order(airport_df$in_degree, decreasing = TRUE),][1:20,c("IATA", "Name", "Country", "City","in_degree")]
top20_in_degree_df## IATA Name Country
## 3483 ATL Hartsfield Jackson Atlanta International Airport United States
## 3631 ORD Chicago O'Hare International Airport United States
## 3171 PEK Beijing Capital International Airport China
## 503 LHR London Heathrow Airport United Kingdom
## 1347 CDG Charles de Gaulle International Airport France
## 3286 LAX Los Angeles International Airport United States
## 337 FRA Frankfurt am Main Airport Germany
## 3471 DFW Dallas Fort Worth International Airport United States
## 3598 JFK John F Kennedy International Airport United States
## 575 AMS Amsterdam Airport Schiphol Netherlands
## 3208 PVG Shanghai Pudong International Airport China
## 3125 SIN Singapore Changi Airport Singapore
## 1187 BCN Barcelona International Airport Spain
## 3552 DEN Denver International Airport United States
## 3726 ICN Incheon International Airport South Korea
## 3377 MIA Miami International Airport United States
## 7630 IST Istanbul Airport Turkey
## 343 MUC Munich Airport Germany
## 2916 HKG Hong Kong International Airport Hong Kong
## 2101 DXB Dubai International Airport United Arab Emirates
## City in_degree
## 3483 Atlanta 911
## 3631 Chicago 550
## 3171 Beijing 534
## 503 London 524
## 1347 Paris 517
## 3286 Los Angeles 498
## 337 Frankfurt 493
## 3471 Dallas-Fort Worth 467
## 3598 New York 455
## 575 Amsterdam 450
## 3208 Shanghai 414
## 3125 Singapore 412
## 1187 Barcelona 392
## 3552 Denver 374
## 3726 Seoul 370
## 3377 Miami 366
## 7630 Istanbul 361
## 343 Munich 360
## 2916 Hong Kong 355
## 2101 Dubai 354top20_out_degree_df <- airport_df[order(airport_df$out_degree, decreasing = TRUE),][1:20,c("IATA", "Name", "Country", "City","out_degree")]
top20_out_degree_df## IATA Name Country
## 3483 ATL Hartsfield Jackson Atlanta International Airport United States
## 3631 ORD Chicago O'Hare International Airport United States
## 3171 PEK Beijing Capital International Airport China
## 503 LHR London Heathrow Airport United Kingdom
## 1347 CDG Charles de Gaulle International Airport France
## 337 FRA Frankfurt am Main Airport Germany
## 3286 LAX Los Angeles International Airport United States
## 3471 DFW Dallas Fort Worth International Airport United States
## 3598 JFK John F Kennedy International Airport United States
## 575 AMS Amsterdam Airport Schiphol Netherlands
## 3208 PVG Shanghai Pudong International Airport China
## 3125 SIN Singapore Changi Airport Singapore
## 1187 BCN Barcelona International Airport Spain
## 3726 ICN Incheon International Airport South Korea
## 343 MUC Munich Airport Germany
## 3377 MIA Miami International Airport United States
## 3552 DEN Denver International Airport United States
## 7630 IST Istanbul Airport Turkey
## 498 LGW London Gatwick Airport United Kingdom
## 2101 DXB Dubai International Airport United Arab Emirates
## City out_degree
## 3483 Atlanta 915
## 3631 Chicago 558
## 3171 Beijing 535
## 503 London 527
## 1347 Paris 524
## 337 Frankfurt 497
## 3286 Los Angeles 492
## 3471 Dallas-Fort Worth 469
## 3598 New York 456
## 575 Amsterdam 453
## 3208 Shanghai 411
## 3125 Singapore 408
## 1187 Barcelona 391
## 3726 Seoul 370
## 343 Munich 368
## 3377 Miami 368
## 3552 Denver 361
## 7630 Istanbul 358
## 498 London 356
## 2101 Dubai 3562.1.3 Degree Histogram & Statistics
degree_hist <- ggplot(degree_df,aes(x= degree)) +
geom_histogram(binwidth = 10, fill = "lightblue") +
xlab(label = "Degree Distribution") +
theme_classic()
degree_hist
psych::describe(degree_df$degree)## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 3425 39.51 106.72 8 14.63 8.9 1 1826 1825 6.03 51.73 1.82We could see we have a very right-skewed distribution, as most of the airports have small number of degree, while the top tiers have plenty.
Who are the medians?
eightdegree_df <- airport_df[which(airport_df$degree==8),c("IATA", "Name", "Country", "City","degree")]
sample_n(eightdegree_df, 20)## IATA Name Country
## 1 ERI Erie International Tom Ridge Field United States
## 2 TMR Aguenar – Hadj Bey Akhamok Airport Algeria
## 3 BQN Rafael Hernandez Airport Puerto Rico
## 4 MUA Munda Airport Solomon Islands
## 5 YSM Fort Smith Airport Canada
## 6 HSL Huslia Airport United States
## 7 BVG Berlevåg Airport Norway
## 8 LWS Lewiston Nez Perce County Airport United States
## 9 UUA Bugulma Airport Russia
## 10 BFN Bram Fischer International Airport South Africa
## 11 CMI University of Illinois Willard Airport United States
## 12 GTO Jalaluddin Airport Indonesia
## 13 IDA Idaho Falls Regional Airport United States
## 14 CUC Camilo Daza International Airport Colombia
## 15 YHU Montréal / Saint-Hubert Airport Canada
## 16 CFR Caen-Carpiquet Airport France
## 17 SMR Simón BolÃvar International Airport Colombia
## 18 LBJ Komodo Airport Indonesia
## 19 YAM Sault Ste Marie Airport Canada
## 20 MZR Mazar I Sharif Airport Afghanistan
## City degree
## 1 Erie 8
## 2 Tamanrasset 8
## 3 Aguadilla 8
## 4 Munda 8
## 5 Fort Smith 8
## 6 Huslia 8
## 7 Berlevag 8
## 8 Lewiston 8
## 9 Bugulma 8
## 10 Bloemfontein 8
## 11 Champaign 8
## 12 Gorontalo 8
## 13 Idaho Falls 8
## 14 Cucuta 8
## 15 Montreal 8
## 16 Caen 8
## 17 Santa Marta 8
## 18 Labuhan Bajo 8
## 19 Sault Sainte Marie 8
## 20 Mazar-i-sharif 8These are mostly regional airport which travel to and from 4 other airports.
2.1.4 In Degree & Out Degree Difference
airport_df$degree_diff <- with(airport_df, out_degree - in_degree) most_outgoing <- airport_df[order(airport_df$degree_diff, decreasing = TRUE),][1:20,]
most_outgoing[,c("IATA", "Name", "Country", "City","in_degree", "out_degree")]## IATA Name
## 2003 JED King Abdulaziz International Airport
## 3367 HOU William P Hobby Airport
## 343 MUC Munich Airport
## 3129 BNE Brisbane International Airport
## 3631 ORD Chicago O'Hare International Airport
## 1318 MRS Marseille Provence Airport
## 1347 CDG Charles de Gaulle International Airport
## 3479 STL St Louis Lambert International Airport
## 3548 MDW Chicago Midway International Airport
## 3168 SYD Sydney Kingsford Smith International Airport
## 3515 IAD Washington Dulles International Airport
## 3659 MSP Minneapolis-St Paul International/Wold-Chamberlain Airport
## 3896 JIB Djibouti-Ambouli Airport
## 474 MAN Manchester Airport
## 1078 SID AmÃlcar Cabral International Airport
## 2093 AUH Abu Dhabi International Airport
## 3147 MEL Melbourne International Airport
## 3260 MCI Kansas City International Airport
## 3332 ADQ Kodiak Airport
## 3518 MKE General Mitchell International Airport
## Country City in_degree out_degree
## 2003 Saudi Arabia Jeddah 183 194
## 3367 United States Houston 70 79
## 343 Germany Munich 360 368
## 3129 Australia Brisbane 144 152
## 3631 United States Chicago 550 558
## 1318 France Marseille 129 136
## 1347 France Paris 517 524
## 3479 United States St. Louis 107 114
## 3548 United States Chicago 132 139
## 3168 Australia Sydney 202 208
## 3515 United States Washington 190 196
## 3659 United States Minneapolis 212 218
## 3896 Djibouti Djibouti 17 23
## 474 United Kingdom Manchester 311 316
## 1078 Cape Verde Amilcar Cabral 15 20
## 2093 United Arab Emirates Abu Dhabi 236 241
## 3147 Australia Melbourne 132 137
## 3260 United States Kansas City 77 82
## 3332 United States Kodiak 6 11
## 3518 United States Milwaukee 60 65out_going_plot <- ggplot(most_outgoing, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=most_outgoing$degree_diff) +
geom_text_repel(data = most_outgoing, (aes(x=Longtitude, y= Latitude, label=Name)), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("Most Out Going Airport")
out_going_plot
most_incoming <- airport_df[order(airport_df$degree_diff, decreasing = FALSE),][1:20,]
most_incoming[,c("IATA", "Name", "Country", "City","in_degree", "out_degree")]## IATA Name Country
## 1937 AKL Auckland International Airport New Zealand
## 2005 MED Prince Mohammad Bin Abdulaziz Airport Saudi Arabia
## 3663 PWM Portland International Jetport Airport United States
## 3552 DEN Denver International Airport United States
## 73 YHZ Halifax / Stanfield International Airport Canada
## 3736 ATH Eleftherios Venizelos International Airport Greece
## 4027 CRW Yeager Airport United States
## 1596 LIS Humberto Delgado Airport (Lisbon Portela Airport) Portugal
## 3264 PHX Phoenix Sky Harbor International Airport United States
## 3286 LAX Los Angeles International Airport United States
## 3678 LAS McCarran International Airport United States
## 3788 PMI Palma De Mallorca Airport Spain
## 983 HRE Robert Gabriel Mugabe International Airport Zimbabwe
## 1239 TLS Toulouse-Blagnac Airport France
## 1300 LYS Lyon Saint-Exupéry Airport France
## 4069 SPI Abraham Lincoln Capital Airport United States
## 499 LCY London City Airport United Kingdom
## 1079 BVC Rabil Airport Cape Verde
## 1083 ADD Addis Ababa Bole International Airport Ethiopia
## 1167 LCA Larnaca International Airport Cyprus
## City in_degree out_degree
## 1937 Auckland 117 96
## 2005 Madinah 59 39
## 3663 Portland 18 2
## 3552 Denver 374 361
## 73 Halifax 52 43
## 3736 Athens 206 197
## 4027 Charleston 15 6
## 1596 Lisbon 221 214
## 3264 Phoenix 257 251
## 3286 Los Angeles 498 492
## 3678 Las Vegas 252 246
## 3788 Palma de Mallorca 277 271
## 983 Harare 31 26
## 1239 Toulouse 83 78
## 1300 Lyon 140 135
## 4069 Springfield 5 0
## 499 London 66 62
## 1079 Boa Vista 16 12
## 1083 Addis Ababa 109 105
## 1167 Larnaca 97 93in_coming_plot <- ggplot(most_incoming, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=abs(most_incoming$degree_diff)) +
geom_text_repel(data = most_incoming, (aes(x=Longtitude, y= Latitude, label=Name)), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("Most In Coming Airport")
in_coming_plot
Interestingly, the two Saudi airports Prince Mohammad Bin Abdulaziz Airport in Madinah and King Abdulaziz International Airport in Jeddah, both ranked high for the degree differences, one for incoming one for out going. It may suggest that a lot of people are visiting Saudi Arabia by entering Madinah and leaving through Jeddah, that’s why more routes are accommodating such needs.
2.2 Betweenness Centrality
top20_betweenness_df <- airport_df[order(airport_df$betweenness, decreasing = TRUE),][1:20,]
top20_betweenness_df[,c("IATA", "Name", "Country", "City","betweenness")]## IATA Name
## 3286 LAX Los Angeles International Airport
## 3575 ANC Ted Stevens Anchorage International Airport
## 1347 CDG Charles de Gaulle International Airport
## 503 LHR London Heathrow Airport
## 3631 ORD Chicago O'Hare International Airport
## 3171 PEK Beijing Capital International Airport
## 2101 DXB Dubai International Airport
## 337 FRA Frankfurt am Main Airport
## 3378 SEA Seattle Tacoma International Airport
## 2437 GRU Guarulhos - Governador André Franco Montoro International Airport
## 3125 SIN Singapore Changi Airport
## 192 YYZ Lester B. Pearson International Airport
## 575 AMS Amsterdam Airport Schiphol
## 3483 ATL Hartsfield Jackson Atlanta International Airport
## 7630 IST Istanbul Airport
## 3168 SYD Sydney Kingsford Smith International Airport
## 3129 BNE Brisbane International Airport
## 3816 DME Domodedovo International Airport
## 3598 JFK John F Kennedy International Airport
## 2182 NRT Narita International Airport
## Country City betweenness
## 3286 United States Los Angeles 1034522.4
## 3575 United States Anchorage 820399.3
## 1347 France Paris 813854.2
## 503 United Kingdom London 702368.6
## 3631 United States Chicago 664992.4
## 3171 China Beijing 651405.4
## 2101 United Arab Emirates Dubai 634412.5
## 337 Germany Frankfurt 587555.3
## 3378 United States Seattle 566562.7
## 2437 Brazil Sao Paulo 521839.4
## 3125 Singapore Singapore 504163.9
## 192 Canada Toronto 482539.9
## 575 Netherlands Amsterdam 460926.9
## 3483 United States Atlanta 447437.6
## 7630 Turkey Istanbul 442873.1
## 3168 Australia Sydney 407827.9
## 3129 Australia Brisbane 392096.6
## 3816 Russia Moscow 377396.6
## 3598 United States New York 375816.7
## 2182 Japan Tokyo 369420.6#set lower bound for label showing
thres <-top20_betweenness_df[20, "betweenness"]
betweenness_plot <- ggplot(airport_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
stat_density2d(aes(fill = ..level.., alpha = I(.3)),
size = 1, bins = 5, data = airport_df,
geom = "polygon") +
geom_point(color="red", alpha = .2, size=airport_df$betweenness/100000) +
world_theme +
geom_text_repel(data = subset(airport_df, betweenness>= thres), aes(x=Longtitude, y= Latitude, label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("By Betweenness")
betweenness_plot
Which airports are Top Betweenness but not Top Degree?
`%nin%` = Negate(`%in%`)
for (i in top20_betweenness_df$Name){
if (i %nin% top20_degree_df$Name){
print(i)
}
}## [1] "Ted Stevens Anchorage International Airport"
## [1] "Seattle Tacoma International Airport"
## [1] "Guarulhos - Governador André Franco Montoro International Airport"
## [1] "Lester B. Pearson International Airport"
## [1] "Sydney Kingsford Smith International Airport"
## [1] "Brisbane International Airport"
## [1] "Domodedovo International Airport"
## [1] "Narita International Airport"2.3 Closeness Centrality
top20_closeness_df <- airport_df[order(airport_df$closeness, decreasing = TRUE),][1:20,]
top20_closeness_df[,c("IATA", "Name", "Country", "City","closeness")]## IATA Name Country
## 3884 YWH Victoria Harbour Seaplane Base Canada
## 4207 CXH Vancouver Harbour Water Aerodrome Canada
## 4771 LPS Lopez Island Airport United States
## 337 FRA Frankfurt am Main Airport Germany
## 1561 VDA Ovda International Airport Israel
## 1347 CDG Charles de Gaulle International Airport France
## 503 LHR London Heathrow Airport United Kingdom
## 2101 DXB Dubai International Airport United Arab Emirates
## 575 AMS Amsterdam Airport Schiphol Netherlands
## 3286 LAX Los Angeles International Airport United States
## 3598 JFK John F Kennedy International Airport United States
## 192 YYZ Lester B. Pearson International Airport Canada
## 7630 IST Istanbul Airport Turkey
## 3631 ORD Chicago O'Hare International Airport United States
## 343 MUC Munich Airport Germany
## 3171 PEK Beijing Capital International Airport China
## 2182 NRT Narita International Airport Japan
## 1515 FCO Leonardo da Vinci–Fiumicino Airport Italy
## 3296 EWR Newark Liberty International Airport United States
## 6828 DOH Hamad International Airport Qatar
## City closeness
## 3884 Victoria 6.673785e-06
## 4207 Vancouver 6.526393e-06
## 4771 Lopez 6.121525e-06
## 337 Frankfurt 5.901794e-06
## 1561 Ovda 5.901550e-06
## 1347 Paris 5.899914e-06
## 503 London 5.898731e-06
## 2101 Dubai 5.895079e-06
## 575 Amsterdam 5.894315e-06
## 3286 Los Angeles 5.892092e-06
## 3598 New York 5.890496e-06
## 192 Toronto 5.886959e-06
## 7630 Istanbul 5.884915e-06
## 3631 Chicago 5.884603e-06
## 343 Munich 5.884118e-06
## 3171 Beijing 5.884084e-06
## 2182 Tokyo 5.881626e-06
## 1515 Rome 5.881592e-06
## 3296 Newark 5.880796e-06
## 6828 Doha 5.880796e-06#set lower bound for label showing
thres <-top20_closeness_df[20, "closeness"]
closeness_plot <- ggplot(airport_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
stat_density2d(aes(fill = ..level.., alpha = I(.3)),
size = 1, bins = 5, data = airport_df,
geom = "polygon") +
geom_point(color="red", alpha = .2, size=airport_df$degree/200) +
world_theme +
geom_text_repel(data = subset(airport_df, closeness >= thres), aes(x=Longtitude, y= Latitude, label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("By Closeness")
closeness_plot
for (i in top20_closeness_df$Name){
if (i %nin% top20_degree_df$Name){
print(i)
}
}## [1] "Victoria Harbour Seaplane Base"
## [1] "Vancouver Harbour Water Aerodrome"
## [1] "Lopez Island Airport"
## [1] "Ovda International Airport"
## [1] "Lester B. Pearson International Airport"
## [1] "Narita International Airport"
## [1] "Leonardo da Vinci–Fiumicino Airport"
## [1] "Newark Liberty International Airport"
## [1] "Hamad International Airport"2.4 Eigenvector Centrality
top20_eigen_df <- airport_df[order(airport_df$eigenvector, decreasing = TRUE),][1:20,]
top20_eigen_df[,c("IATA", "Name", "Country", "City","eigenvector")]## IATA Name Country
## 3483 ATL Hartsfield Jackson Atlanta International Airport United States
## 503 LHR London Heathrow Airport United Kingdom
## 3631 ORD Chicago O'Hare International Airport United States
## 3598 JFK John F Kennedy International Airport United States
## 3286 LAX Los Angeles International Airport United States
## 1347 CDG Charles de Gaulle International Airport France
## 3471 DFW Dallas Fort Worth International Airport United States
## 337 FRA Frankfurt am Main Airport Germany
## 3271 SFO San Francisco International Airport United States
## 192 YYZ Lester B. Pearson International Airport Canada
## 575 AMS Amsterdam Airport Schiphol Netherlands
## 3171 PEK Beijing Capital International Airport China
## 3377 MIA Miami International Airport United States
## 3552 DEN Denver International Airport United States
## 3208 PVG Shanghai Pudong International Airport China
## 3726 ICN Incheon International Airport South Korea
## 2182 NRT Narita International Airport Japan
## 1515 FCO Leonardo da Vinci–Fiumicino Airport Italy
## 1197 MAD Adolfo Suárez Madrid–Barajas Airport Spain
## 3553 PHL Philadelphia International Airport United States
## City eigenvector
## 3483 Atlanta 1.0000000
## 503 London 0.7704645
## 3631 Chicago 0.7442810
## 3598 New York 0.7064476
## 3286 Los Angeles 0.6884858
## 1347 Paris 0.5834824
## 3471 Dallas-Fort Worth 0.5284687
## 337 Frankfurt 0.5272327
## 3271 San Francisco 0.4687981
## 192 Toronto 0.4573527
## 575 Amsterdam 0.4411905
## 3171 Beijing 0.4376281
## 3377 Miami 0.4367697
## 3552 Denver 0.4270043
## 3208 Shanghai 0.4084017
## 3726 Seoul 0.4081232
## 2182 Tokyo 0.3996506
## 1515 Rome 0.3898507
## 1197 Madrid 0.3896220
## 3553 Philadelphia 0.3896203#set lower bound for label showing
thres <-top20_eigen_df[20, "eigenvector"]
eigen_plot <- ggplot(airport_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
stat_density2d(aes(fill = ..level.., alpha = I(.3)),
size = 1, bins = 5, data = airport_df,
geom = "polygon") +
geom_point(color="red", alpha = .2, size=airport_df$eigenvector*10) +
world_theme +
geom_text_repel(data = subset(airport_df, eigenvector >= thres), aes(x=Longtitude, y= Latitude, label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("By Eigenvector")
eigen_plot
for (i in top20_eigen_df$Name){
if (i %nin% top20_degree_df$Name){
print(i)
}
}## [1] "San Francisco International Airport"
## [1] "Lester B. Pearson International Airport"
## [1] "Narita International Airport"
## [1] "Leonardo da Vinci–Fiumicino Airport"
## [1] "Adolfo Suárez Madrid–Barajas Airport"
## [1] "Philadelphia International Airport"3. Comunnity detection
We are using the quicker method fastgreedy, so we will have to remove direction from our graph.
3.1 Sizes of Communities
graph <- as.undirected(g)
graph <- simplify(graph)
fastgreedy_communities <- fastgreedy.community(graph)
V(graph)$community <- fastgreedy_communities$membership
sizes(fastgreedy_communities)## Community sizes
## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
## 732 898 800 183 74 155 52 178 37 12 17 12 19 7 25 15 7 18 9 13
## 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
## 8 6 6 6 22 5 10 8 4 4 23 4 7 3 3 7 3 4 4 3
## 41 42 43 44 45 46 47 48 49
## 3 4 3 2 2 2 2 2 2We have obtained 49 communities, we will explore the biggest 4.
3.2 Initial Plotting
plot(fastgreedy_communities, graph, vertex.label = NA)
#Attaching community id to the airport_df
membership_vec <- membership(fastgreedy_communities)
membership_df <-as.data.frame(as.table(membership_vec))
names(membership_df)[1] <- "id"
names(membership_df)[2] <- "community id"
airport_df <-airport_df %>% left_join(membership_df, by = c("IATA" = "id"))We will then take samples of 20 to see how are these communities formed.
sample_n(airport_df[airport_df$`community id` == 1,], 20)[,c("IATA", "Country","Name")]## IATA Country Name
## 1 OSS Kyrgyzstan Osh Airport
## 2 VAA Finland Vaasa Airport
## 3 PED Czech Republic Pardubice Airport
## 4 HRM Algeria Hassi R'Mel Airport
## 5 ACI Guernsey Alderney Airport
## 6 BJV Turkey Milas Bodrum International Airport
## 7 DOL France Deauville-Saint-Gatien Airport
## 8 ERF Germany Erfurt Airport
## 9 UFA Russia Ufa International Airport
## 10 KAO Finland Kuusamo Airport
## 11 TOB Libya Gamal Abdel Nasser Airport
## 12 AJA France Ajaccio-Napoléon Bonaparte Airport
## 13 TRE United Kingdom Tiree Airport
## 14 TFS Spain Tenerife South Airport
## 15 BLE Sweden Borlange Airport
## 16 LDY United Kingdom City of Derry Airport
## 17 XRY Spain Jerez Airport
## 18 KID Sweden Kristianstad Airport
## 19 KVD Azerbaijan Ganja Airport
## 20 FMM Germany Memmingen Allgau Airportsample_n(airport_df[airport_df$`community id` == 2,], 20)[,c("IATA", "Country","Name")]## IATA Country Name
## 1 KDU Pakistan Skardu Airport
## 2 EKS Russia Shakhtyorsk Airport
## 3 WYA Australia Whyalla Airport
## 4 UTP Thailand U-Tapao International Airport
## 5 DMK Thailand Don Mueang International Airport
## 6 IXD India Allahabad Airport
## 7 IUE Niue Niue International Airport
## 8 BNK Australia Ballina Byron Gateway Airport
## 9 YBP China Yibin Caiba Airport
## 10 AVA China Anshun Huangguoshu Airport
## 11 OBO Japan Tokachi-Obihiro Airport
## 12 TOY Japan Toyama Airport
## 13 TWU Malaysia Tawau Airport
## 14 YNB Saudi Arabia Prince Abdulmohsin Bin Abdulaziz Airport
## 15 JJN China Quanzhou Jinjiang International Airport
## 16 CTU China Chengdu Shuangliu International Airport
## 17 NPL New Zealand New Plymouth Airport
## 18 AUU Australia Aurukun Airport
## 19 RAJ India Rajkot Airport
## 20 FUO China Foshan Shadi Airportsample_n(airport_df[airport_df$`community id` == 3,], 20)[,c("IATA", "Country","Name")]## IATA Country Name
## 1 MEC Ecuador Eloy Alfaro International Airport
## 2 TXK United States Texarkana Regional Webb Field
## 3 EKO United States Elko Regional Airport
## 4 ZOS Chile Cañal Bajo Carlos - Hott Siebert Airport
## 5 BGM United States Greater Binghamton/Edwin A Link field
## 6 IPI Colombia San Luis Airport
## 7 SDQ Dominican Republic Las Américas International Airport
## 8 LAX United States Los Angeles International Airport
## 9 ABI United States Abilene Regional Airport
## 10 PBG United States Plattsburgh International Airport
## 11 PQI United States Northern Maine Regional Airport at Presque Isle
## 12 MKE United States General Mitchell International Airport
## 13 EGE United States Eagle County Regional Airport
## 14 XMS Ecuador Coronel E Carvajal Airport
## 15 ALO United States Waterloo Regional Airport
## 16 PDT United States Eastern Oregon Regional At Pendleton Airport
## 17 SOW United States Show Low Regional Airport
## 18 LCK United States Rickenbacker International Airport
## 19 OAJ United States Albert J Ellis Airport
## 20 SJU Puerto Rico Luis Munoz Marin International Airportsample_n(airport_df[airport_df$`community id` == 4,], 20)[,c("IATA", "Country","Name")]## IATA Country Name
## 1 JNU United States Juneau International Airport
## 2 PIP United States Pilot Point Airport
## 3 HYG United States Hydaburg Seaplane Base
## 4 ANC United States Ted Stevens Anchorage International Airport
## 5 PTU United States Platinum Airport
## 6 MOU United States Mountain Village Airport
## 7 VAK United States Chevak Airport
## 8 TLJ United States Tatalina LRRS Airport
## 9 BTI United States Barter Island LRRS Airport
## 10 ENA United States Kenai Municipal Airport
## 11 CHU United States Chuathbaluk Airport
## 12 ATK United States Atqasuk Edward Burnell Sr Memorial Airport
## 13 KFP United States False Pass Airport
## 14 KWN United States Quinhagak Airport
## 15 HSL United States Huslia Airport
## 16 AIN United States Wainwright Airport
## 17 KWK United States Kwigillingok Airport
## 18 EEK United States Eek Airport
## 19 DRG United States Deering Airport
## 20 BKC United States Buckland Airportcommunity_1 <- airport_df %>% dplyr::filter(`community id` == 1)
community_2 <- airport_df %>% dplyr::filter(`community id` == 2)
community_3 <- airport_df %>% dplyr::filter(`community id` == 3)
community_4 <- airport_df %>% dplyr::filter(`community id` == 4)3.3 Graphing the Communities
community_1_plot <- ggplot(community_1, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=community_1$degree/100) +
ggtitle("Community 1")
community_1_plot
Community 1 is focused on Europe, a bit of Middle East and some coastal part of Africa.
community_2_plot <- ggplot(community_2, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=community_2$degree/100) +
ggtitle("Community 2")
community_2_plot
Community 2 is focused on the Asia-Pacific, Central Asia and Middle East.
community_3_plot <- ggplot(community_3, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=community_3$degree/100) +
ggtitle("Community 3")
community_3_plot
Community 3 is focused in US and some parts of Central & South America.
community_4_plot <- ggplot(community_4, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=community_4$degree/100) +
ggtitle("Community 4")
community_4_plot
Community 4 is mostly centralized in Alaska, with few routes.
4. Insights & Analysis
4.1 Where are these routes flying from and to?
country_origin_df <- airport_df %>% dplyr::select("IATA","Country") %>% dplyr::rename(source.airport=IATA)
df_1 <- merge(x = routes_df, y = country_origin_df, by = "source.airport", all.x = TRUE)
df_1 <- df_1 %>% dplyr::rename(Country_origin=Country)
country_destination_df <- airport_df %>% dplyr::select("IATA","Country") %>% dplyr::rename(destination.airport=IATA)
df_2 <- merge(x = df_1, y = country_destination_df, by = "destination.airport", all.x = TRUE)
df_2 <- df_2 %>% dplyr::rename(Country_destination=Country)
df3 <- df_2 %>% dplyr::count(Country_origin, Country_destination, sort=TRUE)
df3 <- df3 %>%dplyr::rename(number_of_routes=n, source=Country_origin, target=Country_destination)
df3[1:30,]## source target number_of_routes
## 1 United States United States 10518
## 2 China China 6877
## 3 Brazil Brazil 1195
## 4 Canada Canada 1167
## 5 India India 1057
## 6 Russia Russia 964
## 7 Australia Australia 776
## 8 Japan Japan 623
## 9 Indonesia Indonesia 611
## 10 Spain Spain 579
## 11 Mexico Mexico 577
## 12 United Kingdom Spain 518
## 13 Spain United Kingdom 512
## 14 France France 483
## 15 Italy Italy 425
## 16 Mexico United States 373
## 17 United States Mexico 369
## 18 United States Canada 364
## 19 Canada United States 363
## 20 Germany Spain 354
## 21 Spain Germany 353
## 22 United Kingdom United Kingdom 309
## 23 Turkey Turkey 306
## 24 Iran Iran 304
## 25 Norway Norway 302
## 26 Malaysia Malaysia 256
## 27 Philippines Philippines 240
## 28 Greece Greece 235
## 29 Colombia Colombia 233
## 30 Germany Italy 221We could see most popular routes are domestic,and are from countries that are either big geographically or population-wise. Then some of the most popular internal routes are:
- UK to Spain
- Spain to UK
- Mexico to US
4.2 Diameter
Diameter: Which is the longest route?
diameter_routes <- diameter(g, directed = TRUE)
print(paste("The diameter of the route graph is", diameter_routes, ", which means one person can go to", diameter_routes, "cities in one go without repeating the places this person has been."))## [1] "The diameter of the route graph is 14 , which means one person can go to 14 cities in one go without repeating the places this person has been."diameter_stops <- get_diameter(g)
diameter_stops <- as.vector(names(diameter_stops))
diameter_df <- airport_df[match(diameter_stops, airport_df$IATA),]
diameter_df <- diameter_df[complete.cases(diameter_df),]
diameter_df## Airport.ID Name
## 2340 5535 Salluit Airport
## 2318 5504 Ivujivik Airport
## 2319 5506 Akulivik Airport
## 2879 6727 Puvirnituq Airport
## 40 62 La Grande Rivière Airport
## 90 146 Montreal / Pierre Elliott Trudeau International Airport
## 759 1665 Geneva Cointrin International Airport
## 298 609 Copenhagen Kastrup Airport
## 9 9 Kangerlussuaq Airport
## 7 7 Narsarsuaq Airport
## 2275 5442 Qaqortoq Heliport
## 2277 5444 Nanortalik Heliport
## City Country IATA Latitude Longtitude degree in_degree
## 2340 Salluit Canada YZG 62.17940 -75.66720 4 2
## 2318 Ivujivik Canada YIK 62.41730 -77.92530 4 2
## 2319 Akulivik Canada AKV 60.81860 -78.14860 4 2
## 2879 Puvirnituq Canada YPX 60.05060 -77.28690 8 4
## 40 La Grande Riviere Canada YGL 53.62530 -77.70420 6 3
## 90 Montreal Canada YUL 45.47060 -73.74080 371 186
## 759 Geneva Switzerland GVA 46.23810 6.10895 329 163
## 298 Copenhagen Denmark CPH 55.61790 12.65600 457 228
## 9 Sondrestrom Greenland SFJ 67.01222 -50.71160 16 8
## 7 Narssarssuaq Greenland UAK 61.16050 -45.42600 10 5
## 2275 Qaqortoq Greenland JJU 60.71568 -46.02992 14 7
## 2277 Nanortalik Greenland JNN 60.14188 -45.23298 8 4
## out_degree betweenness closeness eigenvector degree_diff community id
## 2340 2 9.583333e+00 5.330803e-06 5.153691e-12 0 5
## 2318 2 4.489352e+03 5.428292e-06 6.067198e-10 0 5
## 2319 2 1.122435e+04 5.529444e-06 1.072068e-07 0 5
## 2879 4 1.797143e+04 5.634438e-06 1.894363e-05 0 5
## 40 3 1.220500e+04 5.741847e-06 1.665524e-03 0 5
## 90 185 3.114321e+05 5.854869e-06 2.926011e-01 -1 3
## 759 166 1.250666e+04 5.850862e-06 1.878552e-01 3 1
## 298 229 3.178263e+05 5.868028e-06 2.566127e-01 1 1
## 9 8 2.369665e+05 5.757417e-06 1.454688e-03 0 7
## 7 5 7.741508e+04 5.654669e-06 1.040234e-05 0 7
## 2275 7 6.062100e+04 5.549359e-06 5.921119e-08 0 7
## 2277 4 2.023300e+04 5.447543e-06 3.388999e-10 0 7diameter_plot <- ggplot(diameter_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .5, size=2) +
geom_text_repel(aes(label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf)x1 <- diameter_df[1,"Longtitude"]
x2 <- diameter_df[2,"Longtitude"]
x3 <- diameter_df[3,"Longtitude"]
x4 <- diameter_df[4,"Longtitude"]
x5 <- diameter_df[5,"Longtitude"]
x6 <- diameter_df[6,"Longtitude"]
x7 <- diameter_df[7,"Longtitude"]
x8 <- diameter_df[8,"Longtitude"]
x9 <- diameter_df[9,"Longtitude"]
y1 <- diameter_df[1,"Latitude"]
y2 <- diameter_df[2,"Latitude"]
y3 <- diameter_df[3,"Latitude"]
y4 <- diameter_df[4,"Latitude"]
y5 <- diameter_df[5,"Latitude"]
y6 <- diameter_df[6,"Latitude"]
y7 <- diameter_df[7,"Latitude"]
y8 <- diameter_df[8,"Latitude"]
y9 <- diameter_df[9,"Latitude"]
diameter_plot +
geom_curve(aes(x = x1, y = y1, xend = x2, yend = y2, colour = "black")) +
geom_curve(aes(x = x2, y = y2, xend = x3, yend = y3, colour = "black")) +
geom_curve(aes(x = x3, y = y3, xend = x4, yend = y4, colour = "black")) +
geom_curve(aes(x = x4, y = y4, xend = x5, yend = y5, colour = "black")) +
geom_curve(aes(x = x5, y = y5, xend = x6, yend = y6, colour = "black")) +
geom_curve(aes(x = x6, y = y6, xend = x7, yend = y7, colour = "black")) +
geom_curve(aes(x = x7, y = y7, xend = x8, yend = y8, colour = "black")) +
geom_curve(aes(x = x8, y = y8, xend = x9, yend = y9, colour = "black")) +
ggtitle("Diameter Path")
4.3 Zoom in on Specific Cities
Where are the the places connected to Madrid
#CREATING SUBFRAMES WITH THE LATITUDE AND LONGITUDE FOR THE AIRPORT OF DEPARTURE AND THE AIRPORT OF ARRIVAL
coords_origin <- airport_df %>% dplyr::select('Airport.ID', 'Latitude', 'Longtitude') %>% dplyr::rename(SourceLat=Latitude,SourceLong=Longtitude)
coords_destiny <- airport_df %>% dplyr::select('Airport.ID', 'Latitude', 'Longtitude') %>% dplyr::rename(DestLat=Latitude, DestLong=Longtitude)
flights_to_from <- routes_df %>%
filter((routes_df$source.airport=="MAD") | (routes_df$destination.airport=="MAD"))
#MERGING WITH ROUTES DATAFRAME BUT USING THE AIRPORT OF DEPARTURE AS COMMON COLUMN
flights_coords_origin <- merge(flights_to_from, coords_origin, by.x='source.airport.id', by.y='Airport.ID' )
#MERGING THE DATAFRAME FLIGHTS_COORDS_ORIGIN WITH THE COORDINATES OF ARRIVAL USING THE AIRPORT OF ARRIVAL AS COMMON COLUMN
flights_coords_destination <- merge(flights_to_from, coords_destiny, by.x='destination.airport.id', by.y='Airport.ID')
flights_with_coords <- merge(flights_coords_origin, flights_coords_destination)
#SUBFRAMING AND EXTRACTING ONLY COORDINATES OF ORIGIN AND DESTINATION
coords <- flights_with_coords %>% dplyr::select('SourceLat','SourceLong','DestLat','DestLong')#CREATING DATAFRAME WITH THE COORDINATES OF ORIGIN
source_df<-data.frame(SourceLong=coords$SourceLong,SourceLat=coords$SourceLat)
#TRANSFORMING THEM INTO SPATIALPOINTS
source_sp<-SpatialPoints(source_df, proj4string=CRS("+proj=longlat"))
#CREATING A DATAFRAME OUT OF THOSE POINTS
source_spdf <- SpatialPointsDataFrame(source_sp, data = source_df)#CREATING DATAFRAME WITH THE COORDINATES OF DESTINATION
dest_df<-data.frame(SourceLong=coords$DestLong,SourceLat=coords$DestLat)
#TRANSFORMING THEM INTO SPATIALPOINTS
dest_sp<-SpatialPoints(dest_df, proj4string=CRS("+proj=longlat"))
#CREATING A DATAFRAME OUT OF THOSE POINTS
dest_spdf <- SpatialPointsDataFrame(dest_sp, data = dest_df)comb_df<-data.frame(coords)
comb_df$distance<-distHaversine(source_sp,dest_sp)#ALLOWS US TO RETRIEVE THE MOST COMMON FLIGHTS BY COORDINATES
source_da <- factor(sprintf("%.2f:%.2f",comb_df[,2], comb_df[,1]))
freq <- sort(table(source_da), decreasing=TRUE)
frequent_destinations <- names(freq)[1:50]
idx <- source_da %in% frequent_destinations
LongLat <- unique(comb_df[idx,1:2])
frequent_flights <-comb_df[idx,]#PLOTTING THE DATAFRAME SO WE GET THE GLOBE
(earth <- system.file("images/world.jpg", package="threejs"))## [1] "/Library/Frameworks/R.framework/Versions/4.0/Resources/library/threejs/images/world.jpg"test_df <- data.frame(origin_lat = comb_df[,1], origin_long = comb_df[,2], dest_lat = comb_df[,3], dest_long = comb_df[,4])
#
globejs(img=earth, lat=LongLat[,1], long=LongLat[,2], arcs=test_df,
arcsHeight=0.3, arcsLwd=2, arcsColor="red", arcsOpacity=0.15,
atmosphere=TRUE,bg="white", height = 800 , width = 800)4.4 Travelling Between 2 Specific Stops
If I’m a consultant based in Madrid and I go to Shanghai every month, which airline should I pick?
distances(g, "MAD", "PVG")## PVG
## MAD 2n_mad <- neighbors(g, "MAD", mode = c('in'))
n_pvg <- neighbors(g, "PVG", mode = c('out'))
middle_stops <- as.table(intersection(n_mad, n_pvg))
names(middle_stops)## [1] "ZRH" "ORD" "BKK" "MUC" "AMS" "ICN" "PEK" "LAX" "DXB" "JFK" "CDG" "FCO"
## [13] "FRA" "HEL" "LHR" "MXP" "SVO" "CPH" "IST" "DOH" "EWR"tempdf <- routes_df %>% dplyr::select("airline","source.airport", "destination.airport") %>% dplyr::rename(source = source.airport, dest = destination.airport)
tempdf1 <- tempdf %>% dplyr::filter(source == "MAD") %>% dplyr::rename(airline1 = airline)
tempdf2 <- tempdf %>% dplyr::filter(dest == "PVG") %>% dplyr::rename(airline2 = airline)sqldf("select tempdf1.*, tempdf2.* from tempdf1, tempdf2 where (tempdf1.dest = tempdf2.source) and airline1 = airline2")## airline1 source dest airline2 source dest
## 1 AA MAD LAX AA LAX PVG
## 2 AA MAD ORD AA ORD PVG
## 3 AF MAD CDG AF CDG PVG
## 4 AY MAD HEL AY HEL PVG
## 5 AZ MAD FCO AZ FCO PVG
## 6 BA MAD LHR BA LHR PVG
## 7 CA MAD PEK CA PEK PVG
## 8 DL MAD JFK DL JFK PVG
## 9 EK MAD DXB EK DXB PVG
## 10 KE MAD ICN KE ICN PVG
## 11 KL MAD AMS KL AMS PVG
## 12 LH MAD FRA LH FRA PVG
## 13 LH MAD MUC LH MUC PVG
## 14 LX MAD ZRH LX ZRH PVG
## 15 MU MAD AMS MU AMS PVG
## 16 QR MAD DOH QR DOH PVG
## 17 SK MAD CPH SK CPH PVG
## 18 SU MAD SVO SU SVO PVG
## 19 TG MAD BKK TG BKK PVG
## 20 TK MAD IST TK IST PVG
## 21 UA MAD EWR UA EWR PVGFrom the list we could see that American Airline(AA) and Lufthansa Airline (LF) are the only two airlines that have more than 1 routes fully operated by themselves. As there could be uncertainty as airports, given more than 1 choice as the pit stop could be better options.
4.5 Adding Passenger Volume
Are the busiest airport really busy? We wanted to add the passenger volume to the data set to evaluate their relation with degree relationship. As the free & available data only has ranked 20 airports, we will do it on a small scale.
passenger_url <- "https://gist.githubusercontent.com/hannahbhchou/01cbc0081c8a080350e50d0ead1a1fcc/raw/33f3a9b29ae6a7323ace128f94775025d23485cb/passenger_2017.csv"
passenger_df <- read.csv(passenger_url, header = TRUE)passenger_df <- passenger_df %>% left_join(airport_df, by = c("IATA" = "IATA"))
passenger_df$v_d_ratio <- with(passenger_df, Volume / degree)passenger_df[,c("IATA", "Name", "Volume", "degree", "v_d_ratio")]## IATA Name Volume degree
## 1 ATL Hartsfield Jackson Atlanta International Airport 103902992 1826
## 2 PEK Beijing Capital International Airport 95786442 1069
## 3 DXB Dubai International Airport 88242099 710
## 4 HND Tokyo Haneda International Airport 85408975 315
## 5 LAX Los Angeles International Airport 84557968 990
## 6 ORD Chicago O'Hare International Airport 79828183 1108
## 7 LHR London Heathrow Airport 78014598 1051
## 8 HKG Hong Kong International Airport 72664075 710
## 9 PVG Shanghai Pudong International Airport 70001237 825
## 10 CDG Charles de Gaulle International Airport 69471442 1041
## 11 AMS Amsterdam Airport Schiphol 68515425 903
## 12 DFW Dallas Fort Worth International Airport 67092194 936
## 13 CAN Guangzhou Baiyun International Airport 65887473 674
## 14 FRA Frankfurt am Main Airport 64500386 990
## 15 IST Istanbul Airport 64119374 719
## 16 DEL Indira Gandhi International Airport 63451503 527
## 17 CGK Soekarno-Hatta International Airport 63015620 367
## 18 SIN Singapore Changi Airport 62220000 820
## 19 ICN Incheon International Airport 62157834 740
## 20 DEN Denver International Airport 61379396 735
## v_d_ratio
## 1 56901.97
## 2 89603.78
## 3 124284.65
## 4 271139.60
## 5 85412.09
## 6 72047.10
## 7 74228.92
## 8 102343.77
## 9 84849.98
## 10 66735.29
## 11 75875.33
## 12 71679.69
## 13 97755.89
## 14 65151.91
## 15 89178.55
## 16 120401.33
## 17 171704.69
## 18 75878.05
## 19 83997.07
## 20 83509.38One thing we’ve noticed that all of the top 20 passenger volume airports are scattered among the most popular communities, but maybe because of their geography they are separated.
passenger_plot <- ggplot(passenger_df, (aes(x = Longtitude, y= Latitude))) +
borders("world", colour=NA, fill="antiquewhite") +
world_theme +
geom_point(color="red", alpha = .2, size=passenger_df$v_d_ratio/18000) +
geom_text_repel( aes(x=Longtitude, y= Latitude, label=Name), color = "black", fontface = "italic", size = 2, max.overlaps = Inf) +
ggtitle("Top 20 Passenger Volume Airports")
passenger_plot
We could see that though Tokyo Haneda airport and Soekarno-Hatta International Airport are the highest in terms of volume/degree ratio, which means for every route they serve more passengers.